What’s New in SAS 9.4 In-Database Products: User’s Guide

Overview

In SAS 9.4, the following new features and enhancements were added to expand the capabilities of the SAS In-Database products:
  • In the January 2016 release of SAS 9.4, several documentation enhancements were made.
  • In the July 2015 release of SAS 9.4, the following changes and enhancements were made:
    • The SAS In-Database Code Accelerator now supports a SET statement with embedded SQL, a SET statement that specifies multiple input tables, and the MERGE statement.
    • In-database processing of PROC TRANSPOSE is preproduction for Teradata and Hadoop.
    • The SAS Scoring Accelerator for Hadoop, SAP HANA, and Teradata supports model scoring using analytic stores.
    • The run and publish model macros for the SAS Scoring Accelerator for Hadoop now support the SAS_HADOOP_CONFIG_PATH environment variable. This eliminates the need for a merged configuration file.
    • The SAS Embedded Process for Hadoop now supports IBM BigInsights, MapR, and Pivotal HD Hadoop distributions.
  • In the February 2015 release of SAS 9.4, the following changes and enhancements were made:
    • The SAS In-Database Code Accelerator for Hadoop uses HCatalog to process complex, non-delimited files. This enables the SAS In-Database Code Accelerator for Hadoop to support Avro, ORC, RCFile, and Parquet file types. HCatalog file formats are not supported on Pivotal HD v2.x or IBM BigInsights v3.x and later.
    • You can now use the DBCREATE_TABLE_OPTS table option to specify the output SerDe, the output delimiter of the Hive table, the output ESCAPED BY character, and any other CREATE TABLE syntax allowed by Hive.
  • In the August 2014 release of SAS 9.4, the following changes and enhancements were made:
    • The SAS Scoring Accelerator and in-database processing of Base SAS procedures is available for SAP HANA.
    • Running limited DATA step scoring programs in Hadoop is now production.
  • In the April 2014 release of SAS 9.4, documentation enhancements were made in the following areas:
    • Considerations when creating or modifying DATA step score code were added.
  • In the December 2013 release of SAS 9.4, the following changes and enhancements were made:
    • Limited DATA step programs can be run inside Hadoop for scoring.
    • New parameters have been added for the Hadoop INDCONN macro variable.
    • The SAS In-Database Code Accelerator for Teradata now can run the DS2 data program as well as the thread program inside the database.
    • DS2ACCEL, a new system option, controls whether the DS2 code is executed inside the database. The default value is NONE, which prevents DS2 code from executing inside the database.
    • The PROC DS2 INDB option has changed its name to DS2ACCEL. INDB is still supported. However, the default value for this option has changed from YES to NO, which prevents DS2 code from executing in the database. This is a change in behavior from the initial 9.4 release.
  • In the September 2013 release of SAS 9.4, the following changes and enhancements were made:
    • In-database processing for Hadoop has been enhanced by the addition of the SAS Scoring Accelerator for Hadoop.
    • The autocall macros that initialized the publishing macros are no longer needed for any DBMS. However, they are still supported.
  • In the July 2013 release of SAS 9.4, the following changes and enhancements were made:
    • The SAS Scoring Accelerator for SPD Server is available.
  • In the June 2013 release of SAS 9.4, the following changes and enhancements were made:
    • Greenplum and Teradata in-database processing has been enhanced by the addition of the SAS In-Database Code Accelerators.
    • The SAS In-Database Code Accelerator enables you to publish a DS2 thread program to the database and execute that thread program in parallel inside the database.
    • Two columns have been added to the model table. The ModelUUID and Notes columns assist in processing scoring models when using the SAS Embedded Process.

SAS In-Database Code Accelerator

July 2015 Release of SAS 9.4: Changes and Enhancements

In the July 2015 release of SAS 9.4, the following changes and enhancements were made:
  • The SET statement can contain embedded SQL.
  • You can specify multiple input tables in the SET statement.
  • The SAS In-Database Code Accelerator for Hadoop supports reading and writing of HDFS-SPD Engine file formats.
  • When a Hadoop data or thread program fails, write a message to the SAS log that contains a link to the MapReduce job log where you can find the error messages.

February 2015 Release of SAS 9.4: Changes and Enhancements

In the February 2015 release of SAS 9.4, the following changes and enhancements were made:
  • The SAS In-Database Code Accelerator for Hadoop supports only Cloudera 5.2 and Hortonworks 2.1 or later. For the latest information, see the SAS Foundation system requirements documentation for your operating environment.
  • The SAS In-Database Code Accelerator for Hadoop uses HCatalog to process complex, non-delimited files.
  • The SAS In-Database Code Accelerator for Hadoop now supports Avro, ORC, RCFile, and Parquet file types.
  • For the SAS In-Database Code Accelerator for Hadoop, you can use the DBCREATE_TABLE_OPTS table option to specify the output SerDe, the output delimiter of the Hive table, the output escaped by, and any other CREATE TABLE syntax allowed by Hive.

August 2014 Release of SAS 9.4: Changes and Enhancements

Hadoop in-database processing has been enhanced by the addition of the SAS In-Database Code Accelerator. The SAS In-Database Code Accelerator enables you to publish a DS2 thread and data program to the database and execute those programs in parallel inside the database.

December 2013 Release of SAS 9.4: Changes and Enhancements

In the December 2013 release of SAS 9.4, the following changes and enhancements were made:
  • The SAS In-Database Code Accelerator for Teradata now runs the DS2 data program as well as the thread program inside the database.
  • DS2ACCEL, a new system option, controls whether the DS2 code is executed inside the database. The default value is NONE, which prevents DS2 code from executing inside the database.
  • The PROC DS2 INDB option has changed its name to DS2ACCEL. INDB is still supported. However, the default value for this option has changed from YES to NO. This change prevents DS2 code from executing in the database. This is a change in behavior from the initial SAS 9.4 release.

SAS 9.4: Changes and Enhancements

The SAS In-Database Code Accelerator enables you to publish a DS2 thread program to the database and execute that thread program in parallel inside the database. Examples of thread programs include large transpositions, computationally complex programs, scoring models, and BY-group processing. The SAS In-Database Code Accelerator is available for Greenplum and Teradata.

Greenplum Changes

SAS 9.4: Changes and Enhancements

You can now specify a non-default port when you create the connection string to publish formats and models.

Hadoop Changes

July 2015 Release of SAS 9.4: Changes and Enhancements

In the July 2015 release of SAS 9.4, the following changes and enhancements were made:
  • The run and publish model macros for the SAS Scoring Accelerator for Hadoop now support the SAS_HADOOP_CONFIG_PATH environment variable. This eliminates the need for a merged configuration file. The INDCONN macro variable has a new argument, and two other arguments are no longer supported.
  • The SAS In-Database Scoring Accelerator for Hadoop now uses HCatalog to process complex, non-delimited files. This enables the SAS In-Database Scoring Accelerator for Hadoop to support all HCatalog file formats, including Avro, ORC, RCFile, and Parquet. HCatalog file formats are not supported on Pivotal HD v2.x or IBM BigInsights v3.x and later.
  • Scoring models using analytic stores is supported.
  • In-database processing of PROC TRANSPOSE is preproduction.
  • The SAS Embedded Process for Hadoop now supports IBM BigInsights, MapR, and Pivotal HD Hadoop distributions.

August 2014 Release of SAS 9.4: Changes and Enhancements

In the August 2014 release of SAS 9.4, the following changes and enhancements were made:
  • You can now specify a fixed record format for the output file of the SAS Scoring Accelerator for Hadoop. Previously, all output was delimited.
  • SPD file formats are supported by the SAS Embedded Process for Hadoop.

December 2013 Release of SAS 9.4: Changes and Enhancements

In the December 2013 release of SAS 9.4, the following changes and enhancements were made:
  • The SAS Embedded Process and the SAS Scoring Accelerator for Hadoop support Kerberos and Hive2 for both Cloudera and Hortonworks.
  • A new parameter, HADOOP_CFG=, is available for the INDCONN macro variable. The HADOOP_CFG= parameter specifies the location of the Hadoop configuration file that is used with the %INDHD_PUBLISH_MODEL and the %INDHD_RUN_MODEL macros.

September 2013 Release of SAS 9.4: Changes and Enhancements

In-database scoring for Hadoop is available.

Netezza Changes

July 2015 Release of SAS 9.4: Changes and Enhancements

If you have Netezza v7.0.3 or later, you can now publish formats and macros to different schemas.

SAP HANA Changes

July 2015 Release of SAS 9.4: Changes and Enhancements

In the July 2015 release of SAS 9.4, the following changes and enhancements were made:
  • Scoring models using analytic stores is supported.
  • Views can be used as input to an SAP HANA scoring model.

August 2014 Release of SAS 9.4: Changes and Enhancements

In-database scoring for SAP HANA is available. You can also run Base SAS procedures inside SAP HANA.

SPD Server Changes

July 2013 Release of SAS 9.4: Changes and Enhancements

In-database scoring for the SAS Scalable Performance Data Server is available.

Teradata Changes

July 2015 Release of SAS 9.4: Changes and Enhancements

In the July 2015 release of SAS 9.4, the following changes and enhancements were made:
  • If you have Teradata v14.10 or later, any object such as column names can be up to 128 characters.
  • In-database processing of PROC TRANSPOSE is preproduction.
  • Scoring models using analytic stores is supported.

DATA Step Processing in Hadoop

August 2014 Release of SAS 9.4: Changes and Enhancements

Running limited DATA step scoring programs in Hadoop is now production.

December 2013 Release of SAS 9.4: Changes and Enhancements

Limited DATA step scoring programs can be run inside Hadoop. This feature is pre-production.

In-Database Procedures

July 2015 Release of SAS 9.4: Changes and Enhancements

In the July 2015 release of SAS 9.4, the following changes and enhancements were made:
  • In-database processing of PROC TRANSPOSE is preproduction for Teradata and Hadoop.
  • In-database processing of Base Procedures is now supported for Impala and HAWQ.

SAS 9.4: Changes and Enhancements

The PRESERVE_NAMES LIBNAME option no longer prevents in-database processing.

Autocall Macros

September 2013 Release of SAS 9.4: Changes and Enhancements

The following autocall macros are no longer needed for any DBMS. However, they are still supported. These macros initialized the publishing macros.
%INDACPF %INDB2PF %INDGPPM %INDNZPM
%INDACPM %INDB2PM %INDNZPC %INDORPM
%INDB2PC %INDGPPC %INDNZPF %INDTDPF
%INDB2PD %INDGPPF %INDNZPJ %INDTDPM

SAS Model Manager Changes

April 2014 Release of SAS 9.4: Changes and Enhancements

A new section was added about considerations when creating or modifying DATA step score code.

SAS 9.4: Changes and Enhancements

Two columns have been added to the model table. The ModelUUID and Notes columns assist in processing scoring models when using the SAS Embedded Process.